learning.sample.uniform: Uniformly Randomly Create a Set of Divisions of 'n' Elements...

In thomasWeise/learnerSelectoR: An R Package for Automatically Selecting and Applying a (Machine) Learning Method

Description

Given a number n of objects, create a list of divisions of these n objects into (larger) training and (smaller) test sets. The resulting list itself contains lists each having two elements, a vector training and a vector test. Both vectors are disjoint (except in case n=1) and together contain all elements from 1:n. The list of lists can then use to perform multiple runs of a learning algorithm, always putting in the training data selected according to the training vector and checking the performance on the data selected according to the test vector.

The idea is that the produced list can be used to perform cross-validation like learning.

Usage

learning.sample.uniform(n)

Arguments

n	the integer number of samples to be drawn

Value

a list of lists, where each list contains two vectors training and test.

thomasWeise/learnerSelectoR documentation built on May 16, 2019, 7:47 a.m.